For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Python function

set_virtual_cpu_target

set_virtual_cpu_target()

max.driver.set_virtual_cpu_target(cpu: str) → None

source

Sets the CPU target for host-independent kernel codegen.

When set before any CPU kernel compilation (e.g. before importing max._interpreter_ops), CPU kernels compile for this fixed target instead of the build host’s CPU, so the kernel cache can ship to and be reused on a different host. Mirrors set_virtual_device_target_arch() for GPUs.

Parameters:

cpu (str) – An LLVM target-CPU name (e.g. “x86-64-v3”, “neoverse-n1”), or “generic” for the most-portable baseline of the host arch family (“x86-64” on x86_64, the armv8-a baseline on AArch64; other families raise an error). Empty string restores host-CPU codegen. “native” is rejected because it would re-leak the build host’s CPU.

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit